Early Prediction Model of Macrosomia Using Machine Learning for Clinical Decision Support.

Area of Science:

• Medical research
• Machine learning applications in healthcare
• Obstetrics and Gynecology

Background:

• Fetal overgrowth (macrosomia) poses significant health risks to mothers and infants.
• Accurate identification of high-risk pregnancies is essential for timely intervention.
• Existing research has gaps in predictive modeling, machine learning integration, and intervention effectiveness for macrosomia.

Purpose of the Study:

• To develop and evaluate a machine learning-based model for predicting macrosomia.
• To address limitations in current research regarding macrosomia prediction and clinical decision-making.
• To leverage maternal characteristics and medical history for improved prediction accuracy.

Main Methods:

• Development of a machine learning model using maternal characteristics and medical history.
• Comparison of three algorithms: logistic regression, support vector machine, and random forest.
• Hyperparameter tuning of the selected logistic regression model using cross-validation.

Main Results:

• The logistic regression algorithm demonstrated superior performance compared to support vector machine and random forest.
• The developed model shows potential for accurately predicting macrosomia.
• Optimized logistic regression model achieved high performance through cross-validation.

Conclusions:

• Machine learning models, specifically logistic regression, can significantly enhance macrosomia prediction.
• Improved prediction facilitates timely interventions in high-risk pregnancies.
• This approach can lead to better health outcomes for mothers and neonates affected by macrosomia.